Identifying General Tumor and Specific Lung Cancer Biomarkers by Transcriptomic Analysis

Abstract

The bioinformatic pipeline previously developed in our research laboratory is used to identify potential general and specific deregulated tumor genes and transcription factors related to the establishment and progression of tumoral diseases, now comparing lung cancer with other two types of cancer. Twenty microarray datasets were selected and analyzed separately to identify hub differentiated expressed genes and compared to identify all the deregulated genes and transcription factors in common between the three types of cancer and those unique to lung cancer. The winning DEGs analysis allowed to identify an important number of TFs deregulated in the majority of microarray datasets, which can become key biomarkers of general tumors and specific to lung cancer. A coexpression network was constructed for every dataset with all deregulated genes associated with lung cancer, according to DAVID's tool enrichment analysis, and transcription factors capable of regulating them, according to oPOSSUM´s tool. Several genes and transcription factors are coexpressed in the networks, suggesting that they could be related to the establishment or progression of the tumoral pathology in any tissue and specifically in the lung. The comparison of the coexpression networks of lung cancer and other types of cancer allowed the identification of common connectivity patterns with deregulated genes and transcription factors correlated to important tumoral processes and signaling pathways that have not been studied yet to experimentally validate their role in lung cancer. The Kaplan-Meier estimator determined the association of thirteen deregulated top winning transcription factors with the survival of lung cancer patients. The coregulatory analysis identified two top winning transcription factors networks related to the regulatory control of gene expression in lung and breast cancer. Our transcriptomic analysis suggests that cancer has an important coregulatory network of transcription factors related to the acquisition of the hallmarks of cancer. Moreover, lung cancer has a group of genes and transcription factors unique to pulmonary tissue that are coexpressed during tumorigenesis and must be studied experimentally to fully understand their role in the pathogenesis within its very complex transcriptomic scenario. Therefore, the downstream bioinformatic analysis developed was able to identify a coregulatory metafirm of cancer in general and specific to lung cancer taking into account the great heterogeneity of the tumoral process at cellular and population levels.

Keywords: breast cancer (BC); coexpression networks; differentially expressed genes (DEGs); early detection and prognosis biomarkers; leukemia (LK); lung cancer (LC).

Figure 1

Number of differentially expressed genes (DEGs) in lung cancer LC and other types of cancer (OTC) microarray studies. Overregulated in blue and downregulated in red. The image was made using GraphPad Prism version 8.00 for Windows, GraphPad Software, La Jolla, California, USA, www.graphpad.com (accessed on 12 June 2019).

Figure 2

Bar graph comparing the fold change expression levels of common winning transcription factors in the three cancer types (LC&BC&LK). Transcription factors with fold change values greater than 1 are upregulated, while those with fold change values less than 1 are downregulated. Inside the boxes are the deregulated transcription factors in most datasets and those that are coexpressed in gene networks. The image was made using GraphPad Prism version 8.00 for Windows, GraphPad Software, La Jolla, California, USA, www.graphpad.com (accessed on 23 June 2022).

Figure 3

Bar graph comparing the fold change expression levels of common winning transcription factors in lung cancer and breast cancer (LC&BC). Transcription factors with fold change values greater than 1 are upregulated, while those with fold change values less than 1 are downregulated. Inside the boxes are the deregulated transcription factors in most datasets and those that are coexpressed in gene networks. The image was made using GraphPad Prism version 8.00 for Windows, GraphPad Software, La Jolla, California, USA, www.graphpad.com (accessed on 23 June 2022).

Figure 4

Bar graph comparing the fold change expression levels of common winning transcription factors in lung cancer and leukemia (LC&LK). Transcription factors with fold change values greater than 1 are upregulated, while those with fold change values less than 1 are downregulated. Inside the boxes are the deregulated transcription factors in most datasets and those that are coexpressed in gene networks. The image was made using GraphPad Prism version 8.00 for Windows, GraphPad Software, La Jolla, California, USA, www.graphpad.com (accessed on 23 June 2022).

Figure 5

Bar graph comparing the fold change expression levels of unique lung cancer winning transcription factors in lung cancer (LC). Transcription factors with fold change values greater than 1 are upregulated, while those with fold change values less than 1 are downregulated. Inside the boxes are the deregulated transcription factors in most datasets and those that are coexpressed in gene networks. The image was made using GraphPad Prism version 8.00 for Windows, GraphPad Software, La Jolla, California, USA, www.graphpad.com (accessed on 23 June 2022).

Figure 6

Biological processes associated with deregulated genes in common between lung cancer and other types of cancer. (A) Downregulated; (B) overregulated. The image was made using GraphPad Prism version 8.00 for Windows, GraphPad Software, La Jolla, California, USA, www.graphpad.com (accessed on 23 June 2019).

Figure 7

Coexpression network of deregulated winning genes common among the three types of cancer (LC&OC coexpression network). In red are the most connected nodes, in orange the average connected nodes, and in yellow the less connected nodes. The TF is highlighted inside the green oval.

Figure 8

Hallmarks of cancer related to the genes of the LC&OC coexpression network.

Figure 9

Coexpression network of unique lung cancer winning deregulated genes that are not deregulated in other types of cancer (LCII coexpression network). In red are the most connected nodes, in orange the average connected nodes, and in yellow the less connected nodes. The TFs are highlighted inside the blue ovals.

Figure 10

Hallmarks of cancer related to the genes of the LCII coexpression network. The TF is in the center of the circle, and the DEGs are more external in the circle. All overregulated genes are in blue, and all downregulated genes are in red.

Figure 11

KM plots of the top winning transcription factors associated with lung cancer patients’ survival, and with a significant logrank p-value. (A) ZBTB16, (B) TAL1, (C) FOXM1, (D) SOX17, (E) EPAS1, (F) KLF2, (G) ID4, (H) MYBL2, (I) NR4A3, (J) FOXF1, (K) GATA6, (L) HOXC6, (M) RFX2.

Figure 12

Venn diagram with the transcriptomic metafirm of winning and coexpressed transcription factors common among the three types of cancer and unique to lung cancer. Negatively regulated transcription factors are in bold red, and positively regulated transcription factors are in bold black.